Lossy vs lossless

[mariush]

I posted some samples here:

http://mplayer.savedonthe.net/test_files/download.php?dir=Samples/

There's 3 versions for every sample, one lossless straight from CD, one "lame -V2" (which should be about 160kbps vbr) and one "lame -preset insane" (which should be mp3 cbr 320kbs). However, the files are all flac and shuffled around for each sample, so you shouldn't be able to guess just by file size which one is.

So feel free to demonstrate you CAN or CAN NOT detect differences between tracks.

If you don't have time to evaluate all 5 samples, I recommend trying at least sample 2 and sample4.

[GodofaGap]

Oh I thought it was some sort of blind test
The point was that you should have done an ABX test on those, not a sighted evaluation. We can ignore your opinions safely for now.

Funnily, most ppl fail/prefer 128kbit to 320kbit: http://www.noiseaddicts.com/2009/03/mp3-sound-quality-test-128-320/
And your link proves this how?

[Ghitulescu]

I actually listened to them several times and each time I thought I could hear a difference such as loss of clarity (I had the player playing them at random) when I looked to see which version was playing it wasn't the version I expected it to be. Several times I thought I'd picked the MP3 and it turned out to be the flac file playing.

You did the same mistake the authors of the article I posted in a previous reply. You (and they) tried to identify the source. That's not the purpose (at least not to me). The purpose is to find whether there is a difference, and since the original is WAV (CD-DA) and the samples sound different, that means the "other one" is somehow falsified.

There are cases in which an MP3 will sound better to some people, because it throws away some bands that contain noise or sounds perceived as noise (ambient noise, tape hiss, vinyl noise etc.). But better or worse are things that are human dependent - some people like the bass, others like the treble, some want to hear the electric guitar and so on. Every change is perceived as better or worse, according to each one's taste.

[yetanotherid]

You did the same mistake the authors of the article I posted in a previous reply. You (and they) tried to identify the source.

No I did not.
I listened for differences but if I thought "this one sounds a little duller" or "this one had more clarity" etc. then I'd look to see which version I was listening to.
Naturally I'd assume the duller one would be the MP3 or the clearer one would be the wave file but it wasn't always the case. Most of the time I couldn't hear a difference anyway, but if I thought I could yet I can't reliably pick which is which... Regardless of why I thought one was the MP3 and one was the wave file.... then how do I know I didn't just imagine the difference?
You're effectively saying that if I can tell the difference between a light blue and a dark blue piece of paper, it's enough to claim I can see a difference even though when I look at one and tell you it's light or dark blue I'm not picking which is which correctly each time. If I can't, then how am I really seeing the difference?

Personally I think the way you perceive sound becomes coloured very quickly (at least mine does) so unless you listen to the same (for example) 4 beats of music back to back and over and over in order to hear differences, by the time you start listening to a different version of the same track your perception has been coloured by the sound of the ending of the previous version. At least when you're comparing samples that sound virtually the same.

Except of course when it comes to people with magic ears such as yourself who can just pick an MP3 off the bat while listening in the car.

Seeing as so far you've not been willing to listen to the samples I uploaded I'm uploading some new ones. Similar to mariush's samples they're all flac files and each was taken from the original wave, CBR MP3 and VBR MP3. You won't even have to think up excuses for not being able to describe the differences to me. All you'll need to do is download the samples and use your magic ears to tell me which one is which.
I'll upload them very shortly.
Maybe we'll get lucky and ramicio will put his ears where is mouth is this time too. ;)

[yetanotherid]

Okay... so here's my new samples. Similar to the last lot they consist of the original audio, a 320kbps CBR MP3 version and a VBR MP3 version. This time however I converted them all back to wave, then converted them all to flac. The only obvious difference between each version is they're numbered.
Included in the sample zip file is another password protected zip file which contains a text file which lists each version and whether it was taken from the original audio or one of the MP3 encodes. Once someone has downloaded the samples, listened to them and posted their results here they can PM me and I'll send them the password for the zip file. That way I can't be accused of tampering with the results, unless of course someone imagines I've tampered with them already, but I can assure you I have not.

I'll look forward to the latest round of excuses offered by ramicio and Ghitulescu for not proving they can easily pick an MP3 as they claim, although to be honest, I'm not sure either of you will be able to top ramicio's previous "ear wax" excuse.

http://ifile.it/8lm7vdg/Music%20Samples.zip

---------------

[MatLz]

@mariush
I must precise my ears are partialy damaged.
Here is my (right or wrong) guess :

sample1
1 cbr 320
2 source
3 vbr V2

sample2
1 vbr V2
2 cbr 320
3 source

sample3
1 cbr 320
2 source
3 vbr V2

sample4
1 vbr V2
2 cbr 320
3 source

sample5
1 vbr V2
2 source
3 cbr 320

Right or wrong, 'source's sound better to my ears :p

[Ghitulescu]

Well, if there's a difference, and the WAV is the original, I assume this is how the composer, the recording engineer, the mixing engineer and the mastering engineer wanted that music to sound. I noticed that recently they don't care too much about the quality, only how loud the music should be (for all young people that use MP3 with not enough power in a noisy environment), and it happens that I noticed that more and more sound engineers revolt against the management and marketing ass-holes and try to revert to the good old sound of the early '90ies.
I see no point to go with a format that alters the original (oh, it offers me a gain in space, but who cares about this today, I don't, maybe the same people that are cheap enough to buy the cheapest player too - once people want to stay in the pig stall and still be comfy and happy, who am I to force them back to standards? and to pay for quality erhh...).
MP3 was a true revolution in the early '90ies, when it gave the possibility of a PC owner to listen "CD-like" music at a fraction of the space (very scarce that days). It also launched the portable revolution. Good old times! Really.
But now, the smallest external HDD can carry 1000 songs in MP3 (2 days of non-stop music) or still 300 in FLAC, a portable can play FLAC and has 2-64GB flashmemory, absolutely no killer advantage, for me. As more and more people noticed that CD-quality is not really CD-quality, they raised the bitrate to the maximum allowed (320) in the hope of maintaining the format (and Fraunhofer to cash some extra money), but still they reached the cul-de-sac. One can very much try to side with horse-carriage manufacturers, it's eco (:)), but they know the times have gone. It's time for you to realise this for MP3, too. It's time to move on.
I side, as always, with the industrial formats (and freeware/open equivalents).

[yetanotherid]

Well, if there's a difference, and the WAV is the original, I assume this is how the composer, the recording engineer, the mixing engineer and the mastering engineer wanted that music to sound.... yade, yada, yada, yada.....

The topic is not storage space. The topic is the ability to pick an mp3 from the original audio. Yours and ramicio's abilities in particular.

I recognize MP3 from WAV on my car, it's neither a silent car (Diesel) nor I have good ears: the trick is to recognize compression artefacts: once you got them you can't overlook them. It happens also on TVs, I can't stand LCDs :)

So download the samples, pick out the MP3s and prove it. It shouldn't be hard and it'll probably take less time than spending the next few hours typing posts full of red-herrings and excuses for not doing so.

[leeperry]

So far there's been nothing but resistance to downloading the current samples from those who claim they can easily pick the difference.
I just woke up, but I'll check them later! OTOH, I won't be able to do a real ABX because my favorite audio player doesn't support that.

I use this thing as low jitter S/PDIF USB transport: http://www.firestone-audio.eu/shop/products/allproductslisted/bravo.php

together w/ this DAC(fit w/ a great sound swappable opamp): http://www.firestone-audio.eu/shop/products/allproductslisted/spitfiremkiipreview.php

They both run off a discrete low ripple PSU: http://www.firestone-audio.eu/shop/products/allproductslisted/supplier220vac.php

and I use heavily modified monitoring grade orthodynamic headphones...these things are very very transparent, I love them.

I don't claim to be able to identify lossy audio at a glance, I just know that MP3 adds THD+N and muddies the SQ...it's all dependent on the source material and the encoding profile being used. Sometimes it's virtually impossible to ID.

[Ghitulescu]

The topic is the ability to pick an mp3 from the original audio.

The topic is lossy vs. lossless. I feel statistically significant differences, you don't. I did answered this once. I don't see the reason to spend more time proving something to you, in particular with a subject I don't particularly like. I am happy with WAV/FLAC, you're happy with MP3. End of [a happy] story.

[yetanotherid]

I don't see the reason to spend more time proving something to you, in particular with a subject I don't particularly like. I am happy with WAV/FLAC, you're happy with MP3. End of [a happy] story.

I can see a reason. Because you're not as confident about being able to easily pick out the MP3s as you claimed and you're frightened you'll prove what most of us already suspect, that you're full of it.

You don't want to spend time proving something to me on a subject you don't really like after having spent days writing post after post on that very subject? And no doubt a subject you'll continue to spend time discussing. Even the excuses are dripping with hypocrisy.
How funny are you? And how obviously full of it? :rolleyes:

[yetanotherid]

Ghitulescu,
While you're furiously wading through your list of excuses for ones you haven't thrashed to death already, how about just for once, answering a simple question asked of you? The way you avoid the ones which don't suit you.... how do you expect anyone to ever take you seriously?

"You're effectively saying that if I can tell the difference between a light blue and a dark blue piece of paper, it's enough to claim I can see a difference even though when I look at one and tell you it's light or dark blue I'm not picking which is which correctly each time. If I can't, then how am I really seeing the difference?"

[Ghitulescu]

You don't get it, do you?

It doesn't matter what I say, what I do, what I prove. You have your fixed ideas and that's it.

I don't care any longer, I'm not Mother Teresa. Be happy with MP3. MP3 is the best. I lied to you, I can't distinguish MP3 from WAV, not with my studio gear. I can't even recognize stereo from mono.

[kypec]

@yetanotherid: what about starting new thread with poll? Inputs would be your concealed samples and each voter marks one of the options for given sample: Source WAV, CBR MP3, VBR MP3?
I think more people could provide their results because now not that many are bothered enough to make posts with comments...

[yetanotherid]

You don't get it, do you?

It doesn't matter what I say, what I do, what I prove. You have your fixed ideas and that's it.

What I do get.... is you can prove what you can prove and my "fixed ideas" (which is kind of funny in itself given I spent a fair bit of time setting up listening tests which you refuse to take) have nothing to do with it.
You claimed you can pick the MP3s, I've uploaded samples which give you the opportunity to show you can in fact do so and my ideas, fixed or otherwise, won't change the results.

I don't care any longer, I'm not Mother Teresa. Be happy with MP3. MP3 is the best. I lied to you, I can't distinguish MP3 from WAV, not with my studio gear.

Fair enough. If you don't care you won't mind if I take your refusal to download the samples and pick out the MP3s as effective proof you can't do so as you claim.
You'll have to explain the Mother Teresa reference because that comes pretty far from left field, even for you.

I can't even recognize stereo from mono.

I think more it'd be more accurate to say you can't recognise fantasy from reality. ;)

[yetanotherid]

@yetanotherid: what about starting new thread with poll?

Yeah, probably a good idea. It's rather lost amongst the endless rhetoric and lame excuses in this thread.

I've got to go out for a bit, I'll look at starting the new thread when I get back. With any luck Ghitulescu won't turn the new thread into a complete mess, given he doesn't even like the topic.

[SeeMoreDigital]

Moderating team: Please close this topic.

[tebasuna51]

Yes, I think all arguments was already exposed (one more time).
Feel free to open a new one, or a poll, to continue.

Thread closed.